Abstract: Tour into the picture (TIP) proposed by Horry et al. is an approach that easily makes animation from one 2D image of a scene. TIP provides a simple 3D scene model, which is only the combination of some billboards and a 3D polygon frame, to generate gratifying high-quality 3D animation. As the key technique of TIP, spidery mesh that bases on the vanishing points in the image provides enough information to do the reconstruction. But the normative relative depth calculating of a single image has not been of attention. In this paper, we propose an approach of relative depth calculating, including viewpoint detection, relative depth calculation (multiple vanishing points), and relative size calculation to get smooth switch between TIP model and the reference image. Without complex calculating, the TIP model can be automatically constructed with respect to the vanishing point(s) specified by the user. We also show that our method of combining TIP with panorama can be applied in representation of cultural heritages practicably and naturally.